As I get older, there have been a lot of things which have been weighing on my mind. But the future has always weighed on my mind. This has been the case at least since I was 10 and we moved to Los Alamos. In fact, it goes back even further, since I was always thinking about what I wanted to be when I grew up: the first time I remember answering that question was when I was 4 on the way back from the beach in Santa Barbara. I wasn’t even asked. I just spontaneously told my parents in the car I wanted to be a doctor. That rapidly switched to paleontologist until I was 10 and my aunt gave me a book on Lucy by Dr White. I didn’t turn out to be a paleontologist at all, but my passion for Deep Time, human origins and all things paleo didn’t fade. But not coincidentally, my interest in the future blossomed at the same time I began to dig into the origins of humanity.
Recently I have been considering the future but in very specific ways. The first is through my involvement in the aerospace industry, GLXP and NASA. This is what my co-blogger calls my attempts to build “yesterday’s tomorrow.” However, I have also been considering the coming to today’s tomorrow: the coming impact of robotics and information technology on the economy. I have been calling it the “Robopocalypse.”
Why did I pick such a hyperbolic name? Well, I was ribbing Noel over his concerns about the future. He is deeply worried the economy and automata. I can attest that I first heard him voice concerns about this back in the antediluvian days of 2006. (Or was it earlier?) I should add that he is not alone.
I thought it was silly. Even if half the current jobs disappeared, it would not happen overnight. Consider: at late as 1940, agriculture employed 19% of the labor force. It took thirty years to push that down to 4.6%. Those were thirty pretty good years, for most people. (It’s now 1.8%.)
However, as I have read and considered more, I do think there will be problems. With a few exceptions, agricultural workers were pulled out by urban opportunities at least as much as they were pushed out by technological advances. The problems we face over 2010-40 will be different and it will take a rather different approach to address them.
Let’s set out the worries.
The Jobs at Risk
This will not be a comprehensive list, but it touch on some obvious problems. We’ll limit it to seven categories.
1. Driving.
This includes taxis, truckers and the like. There are 223,000 taxi drivers and chauffeurs in the United States. There are 1,273,000+ delivery drivers in the U.S. There are 653,300 bus drivers in the U.S. There are 1,700,000+ truckers in the U.S. You are looking at around 4 million people. With the coming of self driving cars, these jobs are all likely to go away.
2. Fast Food
There are 4,438,000+ people working as waiters and servers. 2,148,000 worked as cooks. Of those 4.5 million (+/-) 3.6 million were working in the fast food industry. With the advent of the burgertron and the like, the fast food portion of these jobs is likely to go away.
3. Janitorial
There are 2,324,000 people working as janitors. There are 1,434,000 people working in house cleaning. The bots are coming for their 3.7 million jobs, too.
4. Construction
Construction equipment operators total around 409,700. There are 902,000 carpenters. There are 229,000 masons (brick and concrete). There are 1,284,000 construction laborers. Plus another associated ~600,000 others (1, 2, 3, 4, 5). You’re looking at around 3.5 million jobs — and a lot of them can be automated.
5. Manufacturing
Assembly line workers are about 1,755,000 strong. Machine operators are about 1,013,000. Painting and coating? 149,000. Welders? 357,000. Machinists? 476,000. That’s another 3.75 million jobs.
6. Cashiers
There are 3,338,000 cashiers in the country. Imagine a future where everything in the store has a RFID chip and you get instantaneously charged as you walk out the door. Imagine how National Car Rental today touts it as a benefit that you can rent one of their cars without talking to a person. Imagine the last time you had a pleasant exchange at the 7-11 or Target. Exactly.
7. “Movers” (longshoremen, warehouse workers, etc)
There are 3,428,000 of them.
There you have 25.31 million plus jobs which are in danger and probably will be replacing in the next two decades at the latest. The labor force in the Unit ed States is 156.4 million!
You’re looking at an implosion of at least 1/6th of all jobs in the United States just from those categories. There will likely be more. This is doubly the case since Watson and its buddies (machine learning based software bots) have implications for skilled white collar workers. Especially vulnerable are the assistants and secretaries, as the technology behind Siri and Cortana improves.
You are looking at a potential unemployment not seen since at least the Great Depression assuming these people cannot be trained for other jobs. The crux of the problem is that the jobs they could easily be trained for are, well, gone. Other job categories may expand … but at evenr lower wages.
This is why Noel et al are very nervous about the Robopocalypse. He is concerned about the fact that capital would supplant labor: you can buy a machine with money and it’ll keep working until it breaks, which case you can get another one. That requires a lot of money and what we call the middle class even if we have jobs will not be able to afford it. Never mind the lower or working classes. This would leave us in a virtual state of feudalism. Those with money keep it and get more. Those without ... well. sucks to be us. If this logic is correct, then the best case is that the demand for personal service skyrockets … bringing us back to Downton Abbey but with universal surveillance and weirder clothes.
In fact, if you think of computer programs as a form of capital, then the problem could be even worse. There is a new paper (that I think is wrong, but that will have to wait) that lays out the following logic: once code is written, it doesn’t depreciate. That task is done — it will keep working forever. Great for the person who owns the code; not so good for the people who wrote it. In that world, mandating open-source won’t help. Open-source will keep the owners from gobbling up all the rents but won’t do anything to prevent the wages of coders from crashing.
We can’t let this come about, can we? We certainly can’t let millions starve! And we definitely cannot let a return of feudalism, even techno feudalism, take place.
So, what to do?
There are multiple solutions. They range from the ludditical to utopian (for the silly responses) to even inspired by Milton Friedman of all people (for at least one of the smart ones).
That will be covered in the next Crazy Thoughts Post. But what do you think? Is this a problem for the 2020s? A problem for 2040? Or a problem for never?
Recent Comments